Evidence-based guidelines for primary care of diabetes have been established and disseminated, yet adoption of guidelines in community-based primary care practice has been disappointing. This effectiveness study proposes a randomized trial involving 24 community-based primary care practices to test an innovative intervention to improve diabetes care. The intervention, derived from theoretically based and efficacious programs tested in other settings, adopts a broad focus and seeks to improve diabetes care by a) increasing the practice's organizational capacity to manage change, and b) implementing and sustaining chronic care office systems that support clinician efforts to improve care for diabetes. The intervention will combine two integrated components. The first component will utilize an Improvement Facilitator that will assess the practice's current use of chronic care office systems and their organizational capacity to manage change, provide feedback to key stakeholders in the practice, and work with the practice over six months to form an Improvement Team that will both address organizational capacity to create and sustain improvement and implement chronic care systems for diabetes. In the second component of the intervention, the practice will participate in a local Improvement Collaborative that will afford opportunities to learn and share experiences during implementation and maintenance phases of the intervention with three similar practices in their geographic area. The intervention will be evaluated in two ways. First, a randomized trial using rigorous quantitative methods will measure 2 primary and 3 secondary endpoints at 12 and 24 months, including a) the ADA Physician Recognition Program performance measures by both patient-report and review of the medical record, and b) assessment of the extent to which practices implement and physicians use elements of the chronic care model in their care of diabetes. Change from baseline to 12 months will assess adoption of chronic care improvements, and change from 12 to 24 months will assess sustainability of improvements. Second, a multimethod assessment process will be used to analyze all qualitative and quantitative data separately to understand how and why the intervention led to the observed effects. The practice assessment will strive to understand which components of the interventions were most effective, their relative costs for implementation, and how they might be further improved. Successful components of the intervention will be refined and made available to our collaborators in the project; the Copic Insurance Company and the Colorado Clinical Guidelines Collaborative, for use in their statewide activities to improve diabetes care.